rudder-typer-workflow
Pass
Audited by Gen Agent Trust Hub on Jun 16, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [PROMPT_INJECTION]: The skill processes tracking plan YAML files to generate SDK code, creating an indirect prompt injection surface where malicious content in the tracking plan could attempt to influence the agent's behavior. 1. Ingestion points: Local tracking plan YAML files (e.g., tracking-plan.yaml, properties/*.yaml). 2. Boundary markers: No explicit delimiters or warnings are used in the instructions to separate untrusted data from instructions. 3. Capability inventory: Use of Bash(rudder-cli *) for code generation and workspace management, as well as general file read and write operations. 4. Sanitization: No specific sanitization or validation of the tracking plan content is described beyond the rudder-cli validate command.
- [COMMAND_EXECUTION]: The skill relies on the rudder-cli tool to perform operations such as init, generate, validate, and apply. These commands are standard for the intended workflow of the RudderTyper tool.
- [EXTERNAL_DOWNLOADS]: The CI/CD integration guide references the rudderstack/setup-rudder-cli@v1 GitHub Action. This is an official resource from the vendor and is used for environment setup in automation workflows.
Audit Metadata